Goal-conditioned Reinforcement Learning (G2_balanced) on Walker2d medium v2
1.863Goal-conditioned ReturnPrior Only
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Prior Only2026.04 | 1.863 | 10.604 | |
| PoE2026.04 | 1.838 | 5.72 | |
| KL-Reg2026.04 | 1.836 | 5.774 | |
| Additive2026.04 | 1.784 | 2.655 | |
| Frozen2026.04 | 1.732 | 0 |